Stop finding outages
through customer messages.
Kernus notifies your team when Docker containers crash, restart-loop, or run out of memory. Install in 2 minutes and prevent silent downtime from turning into fire drills.
curl -sSL https://kernus.app/install | shRequires a Unix shell (macOS or Linux Terminal). On Windows, use WSL or Git Bash for this command — or install with PowerShell.
See how a failure
propagates.
Watch in real time how Kernus detects a container crash and alerts your entire team — across every channel, in seconds.
Static flow diagram · Pan, zoom, and read how data moves end to end.
Auto-detects crashes in under 10 seconds
Maps dependencies between all your services
Alerts reach your team on 6 channels at once
Captures exit codes and crash reasons
Click any service node to simulate a crash
Kernus
Monitoring all containers
The expensive part is waiting
It's Friday evening. A customer writes that checkout is failing. You SSH into the server and find a container OOM-killed hours ago. Nobody noticed because monitoring was always "next sprint."
Every incident steals roadmap time
A 2-hour outage often becomes 6-10 engineering hours across debugging, hotfixes, and postmortem work. That's feature work pushed to next week.
The financial hit is bigger than the tool price
If your service handles $200/hour, one 2-hour silent outage can burn $400 in revenue before counting churn and support load.
Customer trust drops every time this repeats
When users discover downtime before your team does, confidence erodes quickly - and renewals get harder even after you fix the issue.
"We kept postponing monitoring. Then a container silently died and our webhook processor was down for hours. We lost a full day recovering and explaining to customers."
— Backend developer, 8-person startup
Setup
Two commands.
That's the whole setup.
No YAML, no exporters, no sidecars. One token, one start command — you're monitoring.
Explore the architecture map — agent, ingest, alerts, and channels in one interactive diagram.
Everything you need.
Nothing you don't.
Container health, smart alerts, terminal UI, public status pages — designed for developers who want to ship, not configure dashboards.
See the full product flow (educational, no account).
Run kernus see for a live dashboard in your terminal
Status — docker-api-1
Identity
Status
Timing
Configuration
Quick Stats
Intentionally not included
Not missing — intentionally excluded to keep Kernus fast and affordable.
Try it now
Build an alert rule.
See it fire.
Pick a condition, set a threshold, choose your channels — then simulate what the notification looks like, including the log excerpt your team will actually receive.
Condition
Fire when CPU stays above threshold for a sustained window.
Trigger when
Notify via
Select a channel to preview the exact notification format.
Rule summary
Alert when cpu_above exceeds 90% for 5 min — notify on Slack.
Click "Simulate alert" to preview
Cost analysis
What is monitoring actually
costing you?
The tool bill is the obvious part. The engineering hours maintaining it are the hidden cost.
Datadog, Grafana Cloud, New Relic, etc.
Alert tuning, dashboard fixes, agent updates, on-call debugging
Monthly savings
$801
Annual ROI
$9,612
Hours freed
96h/yr
Engineering time valued at $75/hr. Tool savings vs. current spend. Actual results vary.
Flat pricing. No per-host fees.
20 hosts on Datadog costs $300/month. On Kernus, it's $29. Same hosts, same core health visibility, fraction of the cost.
Free
Dev solo, hobby projects, learning
Best for validating one environment before rolling out to production
- 1 host, 6 containers
- 10 alerts/day
- Email alerts
- 5min collection interval
- 1-day cloud metrics retention
- Metrics history in dashboard
- 2 team members
- CSV export
- Slack/Discord/Telegram/SMS
Pro
Small teams, real production, 5-10 hosts
Best for product squads that need reliable incident alerts without enterprise pricing
- 5 hosts, 30 containers
- 7-day metrics retention
- 100 alerts/day
- All channels (Slack, Discord, Telegram, SMS, webhook)
- 60s collection interval
- 5 team members
- Email support
- Shareable public status page
- Weekly digest emails
- Export data for CSV
Business
Medium teams, multiple environments, 15-30 hosts
Best for teams running staging + production with stricter uptime expectations
- 30 hosts, 150 containers
- 30-day metrics retention
- 400 alerts/day
- All channels
- 30s collection interval
- 10 team members
- Priority support (< 24h)
- Webhook integrations
- Automated recommendations from metrics
- Shareable public status page
- Weekly digest emails
- Export data for CSV
No surprises: Alerts are never blocked during incidents — rate-limited, not paywalled. Start without a credit card, cancel any time, and migrate gradually alongside your current stack.
Honest comparison
Datadog and Grafana are excellent tools — for teams that need APM, traces, and log aggregation. If you mainly need to prevent silent container downtime, here's the practical cost and time-to-value comparison.
Datadog (20 hosts)
Grafana Cloud
instead of 30+ min
per-host surprises
- Container health
- CPU, memory, network
- Built-in alerts
- Slack / Discord / Telegram
- Zero config
- Zero maintenance
- Log storage
- APM / Traces
- Custom dashboards
- Kubernetes (yet)
Honest trade-off: No logs, no APM, no custom dashboards. If you need those, Datadog is the right call. If you just need your containers healthy — Kernus is 10x cheaper.
What we don't do: Kernus doesn't store logs, doesn't do APM, doesn't build custom dashboards. If you need those, Datadog or Grafana Cloud are the right choice. If you just need to know your containers are alive, healthy, and not eating all your RAM — that's what we built Kernus for.
From our Blog
Technical deep dives on Docker monitoring, container observability, and building infrastructure tools.
Datadog is too expensive: we cut our monitoring bill by 90%
A real-world breakdown of how one team slashed their monitoring costs without sacrificing visibility. Real numbers inside.
How we built a monitoring SaaS with 90% gross margins
The architecture decisions, COGS breakdown, and engineering lessons from building Kernus on Clickhouse.
Docker container monitoring: complete guide 2025
Everything you need to know about monitoring Docker in production. Metrics, alerts, and common pitfalls to avoid.
Common questions
The things engineers ask before committing.
Datadog charges per host, per metric, per integration. That model made sense when monitoring was complex infrastructure. Kernus pre-processes metrics on the agent before sending, so we store far less data — and pass the savings on. 20 hosts on Datadog is ~$300/month. On Kernus, it's $29.
About the builder
Indie-built, shipping for teams who outgrew spreadsheets
Alex Fernando
I'm Alex — I build Kernus end to end: agent, backend, and product UI. The goal is straightforward monitoring for Docker without the week-long setup or the enterprise bill.
If you run a small or mid-sized fleet and want alerts that respect your time, this is the stack I wished existed when I was on call. Feedback welcome; I ship fast.
Monitoring shouldn't be
a project
You shouldn't need a week to set up monitoring.
You shouldn't need to learn a query language to see if a container is healthy.
You shouldn't pay more for monitoring than for the infrastructure itself — at any scale.
Install once. Know when things break. Get back to building.
curl -sSL https://kernus.app/install | shRequires a Unix shell (macOS or Linux Terminal). On Windows, use WSL or Git Bash for this command — or install with PowerShell.
Free for up to 6 containers. No credit card required.